Diet modulates metabolic and hepatic responses to chronic pesticide mixture exposure in mice

28 29 Chronic exposure to pesticide mixtures through diet is common, yet their combined metabolic effects 30 and interactions with dietary factors remain unclear. We identified four pesticides prevalent in human 31 exposure (imazalil, thiabendazole, boscalid, lambda-cyhalothrin) and assessed their combined impacts 32 on hepatic metabolism and metabolic homeostasis using human liver cells and male mice fed standard 33 chow or western diets. We found that the pesticide mixture induced metabolic perturbations in human 34 hepatocytes. In addition, the pesticide mixture altered hepatic gene expression in chow-fed mice and 35 exacerbated western diet-induced glucose intolerance, fasting hyperglycemia, and insulin resistance 36 without affecting body weight or liver steatosis. These findings reveal that dietary context influences 37 the metabolic consequences of pesticide mixtures, highlighting the need to consider nutritional status 38 when evaluating environmental contaminant risks. Our results suggest that pesticide mixtures at 39

doses m ay contribute to metabolic dysregulation, particularly under obesogenic dietary 40 conditions. 41 42

43 Chronic dietary exposure, Glucose homeostasis, Liver metabolism, Pesticide mixture, Metabolic effect 44 45 Highlights 46 - Four common pesticides in mixture disrupt metabolism in liver cells 47 - Dietary exposure to this pesticide mixture alters hepatic gene expression in mice 48 - The pesticide mixture exacerbates WD-induced disruptions in glucose homeostasis 49 - Pesticides and diet interact in producing the metabolic effects of a pesticide mixture 50 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 3

51 Extensive use of p esticides in agriculture has contributed to the contamination of air , soil, 52 water and food, exposing humans to these substances through multiple routes. There is growing 53 evidence that exposure to several classes of pesticides may have adverse health effects in humans , 54 particularly metabolic disorders (Expertise Collective Inserm 2021) . Epidemiological studies have 55 established a link between occupational exposure to pesticides and a higher incidence of obesity and 56 type 2 diabetes (T2D) (Arab and Mostafalou 2023). Dietary pesticide exposure profiles have also been 57 associated with T2D risk in the general po pulation (Rebouillat et al. 2022) . Conversely, higher 58 consumption of organic foods —which typically contain lower levels of pesticide residues —has been 59 associated with reduced risks of metabolic syndrome, obesity, and T2D (Baudry et al. 2018; Kesse-60 Guyot et al. 2017, 2020) . However, although obesity and T2D are key risk factors for metabolic liver 61 diseases such as metabolic dysfunction–associated steatotic liver disease (MASLD), and the liver is the 62 main organ of pollutant metabolism, the link between pesticide exposure and MASLD development is 63 less characterized (Rajak, S. et al. 2022). 64 Mechanistic studies suggest that several pesticides may interfere with pathways involved in 65 metabolic regulation and liver function (Ahmad et al. 2024; Expertise Collective Inserm 2013), notably 66 through interactions with hepatic nuclear receptors that regulate lipid and glucose metabolism, 67 inflammation, and detoxification processes (Capitão et al. 201 7; Fujino et al. 2019; Groswald et al. 68 2023; He et al. 2020; Knebel et al. 2018a, 2018b; Léger et al. 2023; Lichtenstein et al. 2020; Yang et al. 69 2023). Moreover, several in vitro and in vivo studies have reported the pro -oxidative properties of 70 pesticides (Jabłońska-Trypuć 2017; Rives et al. 2020; Wang et al. 2022) . Consistent with this , recent 71 data suggest that exposure to certain organophosphate pesticides is associated with biomarkers of 72 liver injury and function in humans (Li et al. 2022). Experimental studies further indicate that individual 73 pesticides can disrupt overall metabolic homeostasis and promote hepatic steatosis (Arciello et al. 74 2013; Wahlang et al. 2019; Yang and Park 2018). 75 Dietary exposure to multiple pesticide residues is widespread, with consumers chronically 76 exposed to complex mixtures at levels below regulatory limits (European Food Safety Authority (EFSA) 77 et al. 2024 ; Baudry et al. 2021; Castorina et al. 2003) . Yet, the simultaneous presence of multiple 78 pesticides may yield additive or more than additive effects (Wang et al. 2023), which are not predicted 79 by single-compound assessments (Cedergreen 2014; Christen et al. 2014; de Sousa et al. 2014; Roustan 80 et al. 2014). These complex interactions have been mostly reported in in vitro models (Hernández et 81 al. 2013; Schmidt et al. 2021; Tait et al. 2022; Lichtenstein et al. 2020 ; Wang et al. 2023) . In recent 82 years, preclinical studies have also supported the notion that mixed pesticides can alter metabolic 83 homeostasis and liver function (Mesnage et al. 2021); Lukowicz et al. 2018). 84 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 4 Despite these advances, the metabolic consequences of chronic dietary exposure to realistic 85 pesticide mixtures, and their interaction with dietary factors remain unclear. In this study we 86 investigated the effects of chronic dietary exposure to a relevant pesticide cocktail at toxicological 87

values on hepatic metabolism and energy homeostasis, and the potential interactions with 88 dietary factors. 89 Based on exposure profiles identified in the NutriNet Santé cohort (Baudry et al. 2021; 90 Rebouillat et al. 2021, 2022) and in vitro data, we selected four pesticides that induced, when 91 combined, metabolic perturbations in human liver cells. The pesticide mixture was then assessed for 92 its long-term effect in mice. The four pesticides were incorporated in a standard chow diet (CD) or in 93 a western diet (WD) at doses allowing mice to be exposed for 20 weeks to two reference doses: the 94 human acceptable daily intake (ADI; an estimate of the amount that can be ingested on a daily basis 95 over a lifetime without appreciable risk to hum an health) or a tenfold higher dose (10ADI or 1/10 96 NOAEL; one-tenth of the no -observed-adverse-effect level , the highest experimentally determined 97 dose at which no statistically or biologically significant effect has been described), for each of the four 98 individual pesticides. Our murine findings provide evidence that exposition to toxicological reference 99 values of each pesticide in mixture can induce molecular and phenotypic effects in a diet specific 100 manner. 101 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 5

102 1. Chemical and reagents 103 All pesticides used were of 98% purity (Sigma-Aldrich, France). Stock solutions at 50 mM were prepared 104 in >99% dimethyl sulfoxide (DMSO, Sigma -Aldrich, France) and stored at −20°C. The designation ITBL 105 5, 15 or 30 indicates that each pesticide (imazalil [IMZ], thiabendazole [TBZ], boscalid [BSC], lambda-106 cyhalothrin [LCT]) in the mixture was at 5, 15 or 30 µM. LC‒MS grade methanol (MeOH) and acetic acid 107 were purchased from Fisher Scientific (Illkirch, France). Ultrapure water was produced using a Milli-Q 108 system (Millipore, Saint-Quentin en Yvelines, France). 109 110 2. Cell culture and pesticide treatments 111 Immortalized human hepatocytes (IHH) were a generous gift from Professor Bart Staels (Institut e 112 Pasteur, Lille, France; Samanez et al. 2012). Cells were seeded in 6- or 12-well plates precoated with 1 113 g/L porcine gelatin (Sigma-Aldrich) in William's medium(Gibco) (10% decomplemented fetal calf 114 serum, Dutscher), penicillin (100 units/mL, Sigma -Aldrich), and str eptomycin (0.1 mg/mL, Sigma -115 Aldrich), glutamine (4 mM, Sigma -Aldrich), dexamethasone (1 nM, Supelco), and bovine insulin (8.4 116 nM, Sigma-Aldrich) at 37°C with 95% humidity and 5% CO 2. Fifteen hours after seeding, cells were 117 cultured in Dulbecco's Modified Eagle Medium (DMEM) (pyruvate 230 µM, bovine serum albumin 118 [BSA] 1 g/L, penicillin 100 units/mL, and streptomycin 0.1 mg/mL [Sigma-Aldrich], glutamine 4 mM 119 [Sigma-Aldrich], dexamethasone 1 nM [Supelco], Gibco) without serum for 6 hours. Subsequently, cells 120 were cultured in DMEM supplemented with 4% fetal calf serum, human insulin (0.1 to 1 nM) (Sigma-121 Aldrich), and glucose (Sigma-Aldrich) (1 mM to 4 mM) according to the experiments and treated with 122 a 1000-fold concentrated solution of the mixture of the four pesticides (ITBL), in dimethyl sulfox ide 123 (DMSO, Sigma-Aldrich). Controls were cells exposed to 0.1% DMSO. 124 125 3. Cell viability 126 IHH cells were seeded in 12-well plates (1 × 105 cells/well) and treated either for 24 hours, 72 hours or 127 10 days at 5, 15, 30 µM with each pesticide alone. For the 10-day treatment, the medium was changed 128 every 48 hours. At the end of the experiment, the cells were collected upon trypsin treatment (Sigma-129 Aldrich), and were counted using the Luna IITM cell counter after staining with trypan blue (Sigma -130 Aldrich). 131 132 4. Neutral lipid quantification in IHH cells and liver samples 133 IHH cells were seeded in 6-well plates (2.8 × 105 cells/well) and treated every 2 days for 10 days with 134 the pesticide mixture (ITBL) at 30 μM. The agonist of the liver X receptor (LXR), T-0901317 (T0, 30 μM, 135 Sigma-Aldrich), was added 24 hours before cell recovery and was used as a positive control for 136 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 6 intracellular neutral lipid quantification. After 10 days of treatment, cells were harvested (2 wells were 137 pooled) by scraping in EGTA (aqueous solution) 5 mM: methanol (1:2, v/v). For mouse liver samples, 138 the equivalent of 2 mg of tissue was homogenized in FastPrep tubes containing beads and aqueous 139 EGTA 5 mM:methanol (1:2, v/v), then processed using a Precellys, as described previously (Bligh and 140 Dyer, 1959). 141 Lipid extraction was performed by adding 2.5 mL of methanol (Sigma-Aldrich), 2 mL of Milli-Q water, 142 and 2.5 mL of dichloromethane (Fisher Chemical) (2.5:2.5:2, v/v/v) to cell and liver samples. For cells, 143 evaporation was performed twice before resuspending the extracts in 20 μL of ethyl acetate (Sigma -144 Aldrich). For liver samples, a single evaporation was performed before resuspe nding the extracts in 145 160 µL of ethyl acetate. A standard mixture composed of 6 μg of stigmasterol (2 μg/10 μL), 6 μg of 146 cholesterol C17 (2 μg/10 μL), and 16 μg of triglycerides TG19 (4 μg/10 μL) was used. Lipids 147 (triglycerides, free cholesterol, and cholesterol esters) were quantified by gas chromatography coupled 148 with flame ionization detection (GC -FID) (Lipidomics Platform, I2MC, Toulouse), using a Thermo 149 Electron system focused with a Zebron -1 Phenomenex fused silica capillary column (5 m, 0.32 mm 150 internal diameter, 0.50 µm film thickness; Phenomenex, England), as previously described (Podechard 151 et al. 2018) . The oven temperature was programmed to increase from 200 °C to 350°C at a rate of 152 5°C/min, and the carrier gas was hydrogen (0.5 bar). The injector and detector were set at 315°C and 153 345°C, respectively. 154 155 5. Measurement of mitochondrial oxygen consumption rate 156 IHH cells were treated for 24 hours with the pesticide mixture (ITBL) at 30 μM in DMEM medium (XFe24 157 Cell Culture Microplates pre-coated with gelatin, 6.25 × 10 5 cells/well). Then, cells were incubated in 158 Seahorse XF DMEM Medium, pH 7.4 (Agilent) (Seahorse XF Glucose (1 mM final), Seahorse XF Pyruvate 159 (0.1 mM final), and Seahorse XF L -Glutamine (0.2 mM final). Real-time measurements of the oxygen 160 consumption rate (OCR) were performed by isolating a small volume (approximately 5 µL), also known 161 as a "transient microchamber", above the cell monolayer using the Seahorse XF Cell Mito Stress Test 162 (Agilent, Santa Clara, CA, US). 163 Through an integrated drug delivery system, three compounds were sequentially added to the wells 164 (30-minute intervals between each injection): oligomycin (2 μM, ATP synthase inhibitor), carbonyl 165 cyanide-4 (trifluoromethoxy) phenylhydrazone (FCCP, 2 μM, mitochondrial uncoupler), 166 rotenone/antimycin A (0.5 µM each, inhibitors of complex III of the respiratory chain), to determine 167 ATP production, maximal respi ration, and proton leak, respectively. Data were analyzed using 168 Seahorse XFe Wave software (Agilent). The data were normalized to the cell density in each well 169 measured by an automated IncuCyte cellular imaging system (Sartorius). 170 171 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 7 6. Animals and diets 172 In vivo studies were conducted in accordance with EU directive 2010/63/EU for animal experiments 173 and approved by an independent ethics committee under authorization number 17430 -174 2018110611093660. All mice were housed at 21–23°C with a 12h/12h light/dark cycle and had access 175 to standard rodent chow diet (SAFE A04 U8220G10R from SAFE Augy, France) and tap water. Eight -176 week-old male SOPF C57BL/6 mice (Janvier Labs) were acclimatized for one week and then randomly 177 assigned to different experimental groups. Mice were fed either a standard chow diet (CD; 70% 178 carbohydrate, 4% fat, and 14% protein, n = 36) or a Western diet (WD; 61% carbohydrate, 20% fat, and 179 14% protein, n = 36) ad libitum. After 5 weeks, both groups were further divided into 3 subgroups: (i) 180 a group fed a diet containing the mixture of 4 pesticides (ITBL) and exposed to the acceptable daily 181 intake (CD -ADI or WD -ADI) of each pesticide ; (ii) a group fed a diet containing the mixture of 4 182 pesticides (ITBL) and exposed to 10ADI, corresponding to 10 times the ADI (CD-10ADI or WD-10ADI), 183 of each pesticide; and (iii) one group not exposed to pesticides (CD or WD). The exposure period lasted 184 for 20 weeks (n = 12 animals/group). Rodent diets were prepared in collaboration with the SAAJ unit 185 (Jouy-en Josas) as described previously (Lukowicz et al. 2018). The quantities of pesticides incorporated 186 into the rodent diet were confirmed by LC -MS analysis (Eurofins, France) ( supplementary Table 1). 187 Body weight, food intake, and water consumption were monitored weekly throughout the experiment. 188 189 7. RNA extraction of IHH cells and liver samples 190 IHH cells (2 × 105 cells/well, 6 well-palte) and treated for 24 hours with the mixture (ITBL) at 30 µM. 191 The cell monolayers or the liver samples were lysed using TriReagent (MRC). After addition of 192 chloroform (Fisher Chemical), total RNAs were extracted in the aqueous phase and then precipitated 193 with 99.8% isopropanol (Sigma-Aldrich. After washing with 70% ethanol (Sigma-Aldrich), the RNA was 194 resuspended in RNase/DNase -free water (A mbion). The RNA concentration was measured using a 195 nanophotometer (Nanodrop 1000, Thermo Scientific) at an absorbance of 260 nm. The RNA was 196 diluted to a concentration of 135 ng/µl for microarray analysis (IHH cells) or RNA sequencing (liver 197 samples). 198 199 8. Gene expression analysis 200 a. Quantification of relative mRNA expression by RT-qPCR 201 To perform real-time quantitative PCR, 2 µg of RNA was reverse transcribed using a High-Capacity 202 cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA, USA). Amplification reactions 203 were carried out in 96 -well plates in a mixture consisting of SYBR Green (Low ROX SYBR MasterMix 204 dTTP blue, Takyon), a fluorescent DNA intercal ating agent, primer pairs of interest at a final 205 concentration of 300 nM or 900 nM depending on primer efficiency, and cDNA diluted to a 1:20 ratio 206 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 8 in ultrapure DNase/RNase -free water. The primer sequences used are presented in supplementary 207 Table 2. 208 qPCR experiments were performed using an AriaMx real-time PCR system (Agilent). Fluorescence data 209 were analyzed using LinRegPCR software to calculate the PCR efficiency per point and provide a relative 210 initial mRNA concentration through linear regression of the exponential phase of the PCR curve. The 211 relative expression measurements of the genes of interest were normalized to the expression level of 212 the mRNA encoding the GAPDH protein (glyceraldehyde-3-phosphate dehydrogenase). 213 b. Gene expression profiling of IHH cells by microarray 214 Gene expression profiles were obtained for six different cell passages (p26 to p31) on the GeT -TRiX 215 platform (GenoToul, Genopole Toulouse Midi-Pyrénées) using Agilent SurePrint G3 Human GE v3 DNA 216 microarrays ( 8 × 60K, model 072363) follo wing the manufacturer's instructions. Microarray data 217 acquisition was achieved from 200 ng of total RNA as described previously (Lukowicz et al. 2018) . 218 Microarray data and experimental details are available in the Gene Expression Omnibus (GEO) 219 database at NCBI (GSE305353). 220 Microarray data were analyzed using R ( https://www.R-project.org) and Bioconductor packages 221 (Huber et al. 2015) as described previously (Lukowicz et al. 2018) . Enrichment analysis for biological 222 processes in the gene ontology (GO) was performed using Metascape (Zhou et al. 2019) , and 223 transcription factor enrichment was assessed using TRRUST. 224 c. Gene expression profiling of liver samples by RNA sequencing 225 For each of 66 samples, RNA-seq libraries were constructed from 1000 ng of total RNA at the GeT‐TRiX 226 facility (GénoToul, Génopole Toulouse Midi -Pyrénées) using an Illumina Stranded mRNA Prep kit 227 (Illumina, San Diego, CA, USA) following the manufacturer's instructions adapted to produce librar y 228 sizes compatible with paired -end 150-bp read-length sequencing. The libraries were then pooled to 229 equimolar concentrations and transferred to the GeT-PlaGe facility (GénoToul, Génopole Toulouse 230 Midi-Pyrénées) for sequencing into one lane on an Illumina NovaSeq 6000 using a 2 × 150-bp paired-231 end sequencing mode with a NovaSeq 6000 S4 Reagent Kit v1.5. 232 Bioinformatics treatment was executed with Nextflow v23.10.0 -edge (Di Tommaso et al. 2017) and 233 processed using nf-core/rnaseq v3.14.0 (https://doi.org/https://doi.org/10.5281/zenodo.1400710) of 234 the nf -core collection of workflows (Ewels et al., 2020). Reads were aligned to human genome 235

GRCm39 (build GCA_000001635.9, release: 2023 -04). Sequencing data and experimental 236 details are available in NCBI's Gene Expression Omnibus (Edgar et al., 2002) and are accessible through 237 GEO Series accession number (to be provided). 238 Biostatistics analyses were performed under R v4.3.0 (R Core Team, 2023) as previously described 239 (Chousidis et al. 2025). 240 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 9 Clustering results are shown as a heatmap of expression signals, using the MATRIX application (Lippi 241 and Soub ès 2023) , based on differentially expressed genes (p ≤0.05 and fold change >1). Gene 242 Ontology (GO) enrichment analysis of Biological Processes was performed using Metascape (Zhou et 243 al. 2019) with FDR <5%, using default settings. 244 245 9. Blood and tissue samples 246 Blood samples were collected from the submandibular vein into lithium heparin –coated tubes 247 (Sarstedt, Nümbrecht, Germany) throughout the experiment (weeks 12, 16 and at the time of 248 euthanasia). Plasma was isolated by centrifugation (1500 g, 15 min, 4°C) and stored at −80°C. Following 249 animal euthanasia by cervical dislocation, tissue samples were collected, weighed, dissected, and used 250 for histological analyses or frozen in liquid nitrogen and stored at −80°C until further use. 251 252 10. Oral glucose tolerance test (OGTT) and plasma insulin concentration 253 The OGTT was conducted after 23 weeks of diet. Mice were fasted for 6 hours before receiving a 254 glucose solution (2 g/kg body weight) by gavage. Blood glucose levels were measured from the tail 255 vein using an Accu -Check Performa glu cometer (Roche Diabetes Care France, Mylan, France) 30 256 minutes before and 0, 15, 30, 60, 90, and 120 minutes after receiving the glucose solution. For 257 measurements of plasma insulin concentration (see plasma biochemical analyses), 20 µL of blood was 258 drawn from the tip of the tail vein 30 minutes before and 15 minutes after glucose gavage. 259 260 11. Plasma biochemical analyses 261 The plasma insulin concentration was measured using the We -Met platform ( I2MC, Toulouse , 262 France) with an Insulin Mouse Serum Assay HTRF kit (Revvity). During weeks 5, 12, 16, and 23, fasting 263 blood glucose (6 hours of fasting) was measured from a drop of blood taken from the tail vein , using 264 an Accu-Check Performa glucometer (Roche Diabetes Care France, Mylan, France). Plasma samples 265 were analyzed to determine the levels of alanine aminotransferase (ALT), using a Cobas Mira Plus 266 biochemical analyzer (Roche Diagnostics, Indianapolis, IN, USA) (ANEXPLO facility, Toulouse, France). 267 268 12. Histology 269 Paraformaldehyde-fixed, paraffin -embedded liver tissue sections (3 µm) were stained with 270 hematoxylin and eosin (H&E) for histopathological analysis (n = 12 per group) . The stained liver 271 sections were analyzed blindly for steatosis. The histological features were grouped with the steatosis 272 score evaluated according to Akpolat et al. (Akpolat et al. 2005). 273 Paraformaldehyde-fixed, paraffin -embedded pancreas tissue sections ( 4 µm–thick longitudinal 274 sections) were stained with H&E , scanned with a Pannoramic 250 Flash III microscope , and analyzed 275 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 10 in blinded fashion (n = 6 per group) . The number of islets were quantified and the area of each islet 276 was measured with NDP.view software, version 2.9.29 (Hamamatsu). 277 278 13. Analysis of pesticides and their metabolites in urinary samples 279 Urinary samples were collected during 24 h the week before euthanasia and stored at −80°C. Analysis 280 of pesticides and their metabolites are described in supplementary Table 3. 281 282 14. Statistical analysis 283 Statistical analyses were performed using GraphPad Prism for Windows (version 10.2.; GraphPad 284 Software). Data are presented as mean ± SEM. In vitro data were normalized to the total mean of each 285 experiment before being pooled, with two exceptions: intracellular triglyceride measurements where 286 the data were normalized to the number of cells in each condition before pooling; and the heatmap of 287 genes linked to liver steatosis, hepatotoxicity, and nuclear receptor activation, where the data were 288 normalized to control gene expression. For all experiments in IHH cells, effects were assessed with 289 unpaired t-test, excepted for oxygen consumption rate (OCR) the effects were assessed with two-way 290 ANOVA fo llowed by Tuckey’s post -hoc test. For all animal experiments, differential effects were 291 assessed with one-way ANOVA followed by Tuckey's post -hoc test, excepted for body weight survey 292 and the oral glucose tolerance test (OGTT) a two-way ANOVA followed by Tuckey’s post-hoc test was 293 performed. For histology experiments and urinary metabolites analysis differential effects were 294 assessed with Kruskal–Wallis followed by Dunn's multiple comparisons test. 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 11

311 1. Selection of candidate pesticides 312 We established a list of candidate pesticides to be further evaluated for their effects on hepatic 313 metabolism. We first used recent results from a prospective cohort that characterized the dietary 314 exposure profiles to pesticides in a large sample of French adults with variable consumer habits. This 315 study identified six different exposure clusters in regard to estimated dietary exposure to 25 commonly 316 used pesticides (Rebouillat et al. 2021). In the most exposed cluster (“cluster 3”), we selected the 12 317 pesticides with highest estimated exposure (Rebouillat et al. 2021) (Figure 1). We next refined our 318 selection by examining which of those 12 pesticides have a mode of action on their tar get organisms 319 linked to oxidative stress and lipid metabolism (Leroux 2003 ; https://irac-online.org/mode-of-320 action/classification-online/), two key events in the d evelopment and progression of metabolic liver 321 diseases (Friedman et al. 2018) . This second selection led to a list of nine pesticides. Among them, 322 three pesticides that were banned according to the pesticide use regulations in the European Union 323 (EPHY – ANSES & the EU pesticid e database) were excluded. The remaining six pesticides belong to 324 four different classes: a pyrethroid insecticide (lambda -cyhalothrin), a fungicide from the s trobilurin 325 class (azoxystrobin), a fungicide from the carboxamide class (boscalid) , and three azole fungicides 326 (imazalil, thiabendazole, and tebuconazole). Finally, among the three azole pesticides, we examined 327 those that recently showed positive association with T2D risk in the same French NutriNet -Santé 328 cohort (Rebouillat et al. 2022), as MASLD is frequently associated with diabetes (Stefan & Cusi, Lancet 329 Diabetes Endocrinol, 2022) . This led to a final list of five pesticides, including four fungicides 330 (azoxystrobin [AZX], boscalid [BSC], imazalil [IMZ], thiabendazole [TBZ]) and one insecticide (lambda-331 cyhalothrin [LCT]) (Figure 1). 332 In vitro exposure concentrations were derived from the ADI value of each pesticide. Estimated blood 333 concentrations were calculated assuming a 60kg individual with a 5L blood volume, resulting in 334 estimated concentrations in the micromolar range. Accordingly, the four compounds were tested at 5, 335 15 and 30 µM. 336 Our preliminary experiments on IHH cellular viability revealed that IMZ, TBZ, BSC, and LCT did not 337 drastically alter cell viability in response to acute (24 h), or subchronic (72 hours or 10 days) exposure 338 to increasing concentrations of individual pesticides (supplementary Figure 1A-C). AZX induced 339 cytotoxicity at concentrations as low as 5 µM after 72 hours of exposure (supplementary Figure 1B) 340 and was therefore excluded. Subsequent experiments were conducted using the four pesticides at 30 341 µM, the highest non-cytotoxic dose for both acute and chronic exposure in IHH cells. 342 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 12 343 2. Effects of the mixture of 4 pesticides on hepatic metabolism in vitro 344 We next assessed the effects of the 4 selected pesticides in mixture on IHH lipid metabolism and 345 oxidative stress, two key events in the development and progression of obesity -associated hepatic 346 disease (Mardinoglu 2018) . We performed several in vitro assays in IHH cells targeting molecular 347 initiating and key biological events of the adverse outcome pathway (AOP) for liver steatosis (AOP 34, 348 36, 57, 58, 517, 518 https://aopwiki.org) (supplementary Figure 2) (Mellor et al. 2016; Vin ken et al. 349 2017). 350 We first evaluated the activation of the nuclear receptors PPARα, CAR, PXR, LXR and of the 351 transcription factor AhR (the main molecular initiating event triggering liver steatosis (supplementary 352 Figure 2)) by measuring the relative expression of their respective target genes CYP4A11, CYP2B6, 353 CYP3A4, SREBP1c, and CYP1A1 in IHH cells exposed for 24 h to 30 µM of each pesticide in the mixture. 354 IHH cells exposed to the mixture of the 4 pesticides presented with a significantly higher expression of 355 CYP2B6, CYP3A4, SREBP1 and CYP1A1 compared to untreated cells suggesting activation of CAR, PXR 356 LXR and AhR respectively (Figure 2A). 357 Nuclear receptor activation induces changes in gene and protein expression which are considered key 358 biological events in the steatosis AOP. Thus, we used an untargeted microarray approach to examine 359 the whole pattern of IHH gene expression upon pesticide exposure and investigated the differences in 360 gene expression between untreated IHH cells and those exposed to the mixture of the four compounds 361 (Figure 2 B-E). Principal Component Analysis showed a clear discrimination between untreated and 362 pesticide-treated IHH cells ( Figure 2B). In addition, the number of differentially up - and down -363 regulated genes (DEGs) was increased in cells exposed to the pesticide mixture compared to control 364 non-exposed cells (Figure 2C). Hierarchical clustering of DEGs (p1.5, 4205 365 genes) highlighted two clusters with gene expression levels that differed between unexposed IHH cells 366 and those exposed to the pesticide mixture (Figure 2D). Genes from cluster 1 were downregulated in 367 cells exposed to the pesticide mixture and were linked to cell cycle processes and enriched in E2F1/4, 368 MYC, and TP53 target genes. Genes from cluster 2 were upregulated in cells exposed to the pesticide 369 mixture and are mainly targets of SP1, STAT3, TP53, and ATF4 transcription factors. The top related 370 biological functions were “response to nutrient levels ”, “nuclear receptor meta pathway”, “negative 371 regulation of intracellular signal transduction”, and “response to endoplasmic reticulum stress” (Figure 372 2E). We then focused our analysis on relevant genes involved in liver steatosis, nuclear receptor 373 activation, and hepatotoxicity (Lichtenstein et al. 2020) . As shown in supplementary Figure 3, IHH 374 exposure to the pesticide mixture led to significant upregulation of a large number of these genes, 375 relative to gene expression in untreated cells . The most strongly induced genes are involved in 376 xenobiotic metabolism (SULT1C2, CYP3A5, UGT2B7, CYP2B6, POR) and in lipid metabolism ( PNPLA3, 377 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 13 SCD1, SREBF1, MSMO1, PPARa, HADHB), supporting the potential pro-steatotic impact of the pesticide 378 mixture. 379 We next measured intracellular triglyceride levels in IHH cells treated for 10 days with the pesticide 380 mixture using GC-FID as lipid accumulation is a key event in the AOP for liver steatosis and a hallmark 381 of the disease. Exposure of IHH cells to the pesticide mixture at 30 µM resulted in a higher content of 382 triglycerides compared with that of untreated cells (Figure 2F). 383 At the organelle level, mitochondrial disruption has been proposed to be a late key event in the 384 steatosis AOP. Thus, we next evaluated the effect of combined pesticides on IHH cell mitochondrial 385 respiratory functions using Seahorse XF stress test technology. Exposure of IHH cells to the pesticide 386 mixture led to an increased basal and maximal mitochondrial respiration (Figure 2G) and a significant 387 rise in ATP production (Figure 2H). 388 Altogether, combining data from previously published epidemiological studies and a panel of in vitro 389 assays, we identified four commonly used pesticides that , when combined, induced metabolic 390 perturbations in liver cells. 391 392 3. Chronic dietary exposure to the mixture of 4 pesticides in mice 393 We next investigated the in vivo metabolic and hepatic effects of the pesticide mixture and the 394 interactions with dietary factors. Adult male mice were first fed either a control diet (CD) or a western 395 diet (WD) for 5 weeks and then exposed to the pesticide mixture through these diets for an additional 396 20 weeks (Figure 3A). Pesticides were incorporated in the CD and WD at doses exposing mice to the 397 ADI (CD- or WD-ADI) or 10 times ADI (CD- or WD-10ADI) of each of the four pesticides in the mixture 398 (Figure 3A). Pesticide levels quantified in the feed pellets confirmed that the concentration of the four 399 pesticides in each diet was close to the expected quantities (supplementary Table 1). 400 Body weight did not show any significant differences between exposed (ADI and 10ADI) and non -401 exposed mice in both the CD- and the WD-fed groups (Figure 3B, C). Perigonadal (WATpg) and 402 subcutaneous (WATsc) white adipose tissue weights were also not signific antly changed by pesticide 403 mixture exposure in mice fed a CD or WD, except for a small increase in the relative WATsc weight in 404 the CD-10ADI compared with that in the CD mice (Figure 3D-G). Food and water intake also did not 405 differ between exposed and unex posed mice in both the CD- and the WD-fed mice ( supplementary 406 Figure 4A, B). 407 We next evaluated the impact of pesticide mixture exposure on liver homeostasis during CD and WD 408 feeding. Animal exposure to the pesticide mixture did not impact liver weight nor induce hepatic 409 damage, whatever the dose and the type of diet (CD or WD) (Figure 4A-D). Histological analysis of H&E-410 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 14 stained liver slices and hepatic triglyceride quantification confirmed the absence of pesticide impact 411 on hepatic steatosis in WD-fed mice (Figure 4E, F). By contrast, a slight but significant increase in the 412 steatosis score and a trend toward higher hepatic triglyceride levels were observed in CD-ADI–fed mice 413 compared with unexposed mice (Figure 4E, F). 414 To further explore the potential impact of pesticide exposure on the liver, we analyzed the hepatic 415 transcriptome in each animal group using RNA sequencing. PCA of gene expression profiles showed a 416 slight separation between exposed and non -exposed mice fed a CD along the first principal 417 component, accounting for 19.4% of the variance ( Figure 4G). By contrast, PCA did not allow 418 discrimination between exposed and non -exposed WD -fed mice , whatever the dose of pesticide 419 mixture ( Figure 4H). The number of differentially up - and down -regulated genes in CD -fed mice 420 exposed to the ADI of each pesticide in the mixture (607 up- and 611 down-regulated genes) and to 421 10ADI of each pesticide in mixture (904 up- and 994 down-regulated genes) was higher than it was in 422 unexposed animals (Figure 4I). By contrast, exposure to the pesticide mixture in WD-fed mice did not 423 affect the number of DEGs compared with that in the unexposed animals. We performed hierarchical 424 clustering of DEGs (p1; 2311 genes) in the CD, CD-ADI, and CD-10ADI groups 425 (Figure 4J). Two clusters of genes were identified. Genes from clusters 1 and 2 were respectively up- 426 and down-regulated in exposed CD-fed (CD-ADI and CD -10ADI) compared with their expression in 427 unexposed CD-fed mice (Figure 4J). Upregulated genes are linked to fatty acid metabolism, amino acid 428 metabolism, and cellular respiration, and are mainly enriched in PPARα targets. Downregulated genes 429 from cluster 2 are mainly involved in RNA and protein processing (Figure 4K). 430 Altogether, liver analysis showed that exposure to the pesticide mixture did not exacerbate WD -431 induced alterations in hepatic phenotype and gene expression. However, exposure to the pesticide 432 mixture in CD-fed mice was associated with significant changes in hepatic gene expression. 433 To further investigate the differential hepatic impact of pesticide exposure according to the type of 434 diet, we compared pesticide metabolism in CD- and WD-fed mice. Analyses of urine samples by UHPLC-435 HRMS allowed the detection of several pesticides and their metabolites. As shown in Figure 5A-D and 436 supplementary Table 3, TBZ and BSC and 3 of their metabolites were detected , including phase 1 437 metabolites (boscalid 5-hydroxy, thiabendazole 5-hydroxy) and phase 2 metabolites (glucuronide and 438 sulfate conjugates) . As expected, a ll pesticide metabolites were found at higher levels in urine of 439 animals exposed at 10 times the ADI than in urine from animals exposed to the ADI. Overall, more 440 pesticide metabolites or higher levels were detected in exposed CD-fed mice than in exposed WD-fed 441 mice (figure 5A-D). 442 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 15 To evaluate the pesticide-detoxifying capacities of CD- and WD-fed mice, we then focused the hepatic 443 RNA sequencing analysis on gene s involved in xenobiotic metabolism. The expression profiles of 444 hepatic genes encoding xenobiotic metabolism enzymes in the six experimental animal groups are 445 presented in Figure 5E, F. They reveal that exposure to the pesticide mixture significantly increased 446 the expression of 7 genes in mice fed a CD compared with only 2 in mice fed a WD. 447 Taken together, these results suggest differences in the pharmacokinetics of pesticides, especially in 448 metabolism, between exposed CD- and WD-fed mice. To determine whether extrahepatic pesticide 449 metabolism occurs in WD -fed mice —for example in adipose tissue , which can store lipophilic 450 pollutants—we analyzed WAT for the expression of genes encoding enzymes that regulate 451 detoxification, and of other genes involved in lipolysis, adipogenesis, glucos e metabolism , and 452 inflammation (supplementary Figure 5). None of these gene s’ expression was significantly impacted 453 by pesticide exposure in CD- or WD-fed mice. Similarly, brown adipose tissue (BAT) gene expression of 454 BAT markers and batokines did not significantly differ between pesticide -exposed and unexposed 455 mice, both under WD and CD ( supplementary Figure 6). We also evaluated the impact of pesticide 456 mixture exposure on the digestive tract as the first target of dietary pollutants. Expression analysis of 457 genes involved in the structural integrity and permeability of the ileum of the intestine revealed that 458 males fed a CD and exposed to 10 times the ADI of pesticides in a mixture had reduced expression of 459 several genes involved in the ER stress response (Xbp1s), antimicrobial activity (Reg3b and Reg3g), and 460 permeability (Cldn2) (supplementary Figure 7). However, the expression of none of these gene s was 461 significantly affected by pesticide exposure in WD -fed mice. Together, these results indicate diet -462 dependent differences in pesticide metabolism, suggesting that the bioavailability of pesticides and/or 463 the animals’ detoxifying capacity differ according to the nutritional context. 464 We next evaluated the consequences of chronic exposure to the pesticide mixture on glucose 465 homeostasis. At week 18 of exposure, glucose tolerance, fasting glycemia and insulinemia, and HOMA-466 IR were not affected by pesticide mixture exposure in mice fed a CD (Figure 6A-D). In contrast, WD-fed 467 mice exposed to the pesticide cocktail at 10ADI exhibited significantly higher glucose intolerance, 468 fasting glycemia, insulinemia, and HOMA-IR compared with animals fed the WD but unexposed (Figure 469 6E-H). To further investigate pesticide mixture–induced glucose homeostasis perturbations in WD-fed 470 mice, we analyzed pancreatic endocrine mass and islet number. Although both parameters were 471 significantly higher in the WD-fed mice than in the CD-fed mice , they did not differ significantly 472 between exposed and unexposed animals under WD (Figure 6I, J). Overall, these results show that 473 dietary exposure to the mixture of pesticides amplified WD-induced gl ucose homeostasis 474 perturbations that were not associated with a compensatory increase in pancreatic endocrine mass. 475 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 16

476 In this work, we aimed to evaluate the metabolic consequences of a realistic dietary exposure 477 to a mixture of pesticides at their regulatory reference doses, and to determine whether these effects 478 vary depending on the diet composition. The innovative aspects of our study lies in (i) the integration 479 of human epidemiological exposure profiles to guide our pesticide selection (Baudry et al. 2021; 480 Rebouillat et al. 2021), (ii) the use of several in vitro assays to identify a pesticide mixture that impact 481 metabolic processes in liver cells , (iii) an in vivo study assessing the pertinence of the toxicological 482

values (the ADI and 10ADI; corresponding to 1/10 NOAEL ) of the 4 selected pesticides in 483 mixture and the influence of diet composition on pesticide mixture-induced effects. 484 The aim of our in vi tro strategy was to characterize the metabolic effects of the pesticide 485 mixture in liver cells . Our main finding s were that the selected pesticide mixture induced gene 486 expression changes, triglyceride accumulation and mitochondrial activity perturbation s in IHH cells 487 suggesting its pro-steatotic potency. 488 Our in vivo study revealed that the metabolic effects of the pesticide mixture in mice were not 489 uniform, but rather dependent on the type of diet (CD vs WD) , highlighting the role of nutritional 490 context in shaping toxicological outcomes. While dietary exposure to the pesticide mixture did not 491 elicit significant alterations in the body weight of mice fed either a CD or a WD, it led to a slight but 492 significant increase in the steatosis score in CD -fed mice. The nonsignificant changes in hepatic 493 triglyceride levels in CD fed mice exposed to the pesticide mixture is not entirely consistent with our 494 in vitro studies, which showed a significant increase in triglyceride content in IHH cells upon exposure 495 to the pesticide mixture. However, the changes observed in the hepatic gene expression profile of mice 496 fed a CD support a pro steatotic property of the pesticide mixture. Pathway enrichment analysis of the 497 hepatic transcriptome of CD -fed mice identified fatty acid metabolic processes as the top biological 498 function associated with up-regulated genes in response to pesticide mixture exposure. It cannot be 499 excluded that the 20-week duration of pesticide exposure in our in vivo experiment was insufficient to 500 induce detectable phenotypic alterations. In our previous study, we demonstrated that pesticides 501 induced steatosis after 6 months of exposure (Lukowicz et al. 2018). This observation is consistent with 502 other findings that reported pesticide-induced liver damage in mice fed a control diet over extended 503 periods (Dinca et al. 2023; Docea et al. 2018, 2019; Fountoucidou et al. 2019) . The discrepancy 504 between in vivo and in vitro results may stem from interspecies differences or/and from the inability 505 of in vitro models to fully capture the complex whole-organism liver responses shaped by interorgan 506 interactions, including those with adipose tissue and the gut microbiota (Djekkoun et al. 2021; Nichols 507 et al. 2024; Velmurugan et al. 2017; Wang et al. 2021). 508 While dietary exposure to the pesticide mixture of WD -fed mice did not induce liver 509 phenotypic and genomic alterations, it exacerbated WD-induced diabetic symptoms, including fasting 510 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 17 hyperglycemia, glucose intolerance, and insulin resistance at tenfold ADI dose. These data suggest that 511 exposure to this pesticide mixture may contribute to the development of glucose metabolism 512 disruption in t he presence of other risk factors, such as chronic consumption of high -fat foods. Our 513 findings agree with recent studies showing that high -fat dietary intake may enhance the metabolic 514 effects of pesticide exposure. Oral administration of cypermethrin to adult male mice disrupts glucose 515 homeostasis and induces prediabetic symptoms in high-fat diet-fed animals (Wei et al. 2023a). Other 516 studies in rodents have demonstrated that the metabolic perturbations induced by a high-fat diet are 517 potentiated by exposure to low doses of permethrin, chlorpyrifos, perfluorooctanoic acid, and 518 bisphenol A (Attema et al. 2022; Li et al. 2023; Ma et al. 2021; Wang et al. 2021; Xiao et al. 2018). The 519 interaction between dietary factors and pesticide exposure has primarily been demonstrated for 520 individual compounds, with effects varying based on the specific contaminant and exposure duration. 521 However, studies investigating the interplay between diet and pesticide mixtures remain scarce. In a 522 zebrafish model, exposure to a mixture of organochlorine pesticides ex acerbated the diabetogenic 523 consequences of a high -fat diet (Lee et al. 2023) . Our study extends these findings in mammals, 524 showing that a pesticide mixture can exacerbate WD-induced disruption of glucose homeostasis. 525 Unlike their expression levels in CD-fed mice, the expression levels of xenobiotic metabolizing 526 enzymes were not increased in livers of WD -fed mice following exposure to the pesticide mixture, 527 suggesting reduced pesticide metabolism or the occurrence of extrahepatic pesticide metabolism in 528 WD-fed mice. As urinary profiles of pesticide m etabolites were similar between non -exposed and 529 pesticide-exposed WD-fed mice, we hypothesize that pesticides are overall less metabolized in WD -530 fed mice and may accumulate in other tissues. Previous studies reported that several pesticides , 531 because of their lipophilicity, target adipose tissue (Barrios-Rodríguez et al. 2021; Chang et al. 2016; 532 Jackson et al. 2017; Sousa et al. 2023) . Complementary experiments would be necessary to fully 533 elucidate the fate of pesticides in WD-fed mice. 534 The observed hyperglycemia and insulin resistance in WD -fed mice exposed to the pesticide 535 mixture may not be attributed to alterations in gluconeogenesis or glycogen synthesis or decreased 536 glucose uptake in insulin -sensitive tissues, as observed in other studies (Wei et al. 2023b) . Indeed, 537 while we did not directly measure hepatic glucose uptake, we found that exposure to the pesticide 538 mixture did not affect the expression of genes involved in glucose synthesis and transport or in insulin 539 signaling in the liver (results not shown). This suggests that the liver may not be the primary target 540 tissue through which the pe sticide mixture influences glucose metabolism in WD -fed mice. Whether 541 glucose uptake by skeletal muscle and/or adipose tissue is impaired upon pesticide mixture exposure 542 remains to be determined. 543 T2D occurs when β cells fail to adequately increase insulin secretion to meet demands to 544 counteract insulin resistance, and this failure may be exacerbated by a reduction in β-cell mass over 545 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 18 time (Costes et al. 2021). Several pieces of evidence indicated that pancreatic β cells may be targeted 546 by pollutants (Hectors et al. 2011; Hoyeck et al. 2022; Lee et al. 2017) . In this study, fasting 547 hyperglycemia and increased fasting insulin in response to WD were both higher in mice exposed to 548 pesticide (10ADI or 1/10 NOAEL) than in non-exposed mice, suggesting that the increased secretory 549 function of pancreatic islet β cells was not sufficient to compensate for insulin resistance. In addition, 550 WD-fed mice exposed to the pesticide mixture showed no differences in islet number and endocrine 551 mass when compared with unexposed mice. While we cannot exclude a difference between the 552 numbers of alpha and beta cells, these data suggest that in WD -fed mice exposed to the pesticide 553 mixture, the endocrine pancreas fails to counteract pesticide-induced glucose intolerance and insulin 554 resistance, in contrast to what occurred in males fed a WD but not exposed. 555 In conclusion, our study demonstrates that chronic dietary exposure to reference doses of 556 each pesticide of this realistic cocktail is associated with significant changes in hepatic gene expression 557 in CD-fed mice and exacerbates WD-induced disruption of glucose homeostasis. This is one of the few 558 studies to demonstrate that dietary con text significantly alters the hepatic transcriptomic and 559 metabolic responses to a pesticide mixture in a diet specific manner. The differential response 560 observed between CD- and WD-fed mice emphasizes the complex interplay between environmental 561 contaminants and dietary factors and the importance of considering dietary context when evaluating 562 the metabolic effects of pesticide mixtures. Despite the limitations in translating findings from mice to 563 humans, our results suggest that sensitivity to pesticide expo sure may differ according to metabolic 564 status. Given that toxicological reference values ensuring consumer safety are defined for individual 565 pesticides, our findings suggest that their relevance may differ when pesticides are combined in 566 mixtures. 567 568 569 570 571 572 573 574 575 576 577 578 579 580 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 19

of the study 581 Although our study assessed the effects of a pesticide mixture administered through food intake, at 582 two toxicological reference doses, on 12 animals per group and in two nutritional contexts, it has some 583 limitations. First, we did not identify the mechanisms underlying the effects of the pesticide mixture 584 on glucose homeostasis or the specific tissue in which pesticides accumulate in WD-fed mice. However, 585 our study provides a comprehensive overview of the observed transcriptomi c and phenotypic 586 outcomes, which can serve as a basis for future mechanistic studies. Further mouse studies comparing 587 the effects of the pesticide mixture with those of individual pesticides would be necessary to provide 588 a better understanding of the inter actions among the compounds in the mixture. Finally, we did not 589 evaluate the effects of pesticide exposure in female mice. As energy and xenobiotic metabolism in the 590 liver are highly sexually dimorphic, it is likely that the pesticide cocktail could have sex-specific health 591 effects. 592 593 Acknowledgments 594 This work was supported by the French Foundation for the Medical Research FRM 595 (ENV202109013962), the Caisse Centrale de la Mutualité Sociale Agricole (AAP Mutualité Sociale 596 Agricole MSA 2022-BIOMEC), the department AlimH of INRAE. We thank Professor Bart Staels and Dr 597 N. Hennuyer (Institut Pasteur, Lille, France) for their generous gift of IHH cells. We thank the EZOP 598 staff, the GeT -Trix Genotoul facility, Metatoul -Metabohub, Anexplo, and We-Met facilities for their 599 help. We also thank the INRAE SAAJ –RAF team (Jouy-en-Josas, France) for its technical support with 600 the pellet preparation. C.R. was supported by FRM (ENV202109013962). 601 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 20

602 Ahmad MF, Ahmad FA, Alsayegh AA, Zeyaullah Md, AlShahrani AM, Muzammil K, et al. 2024. 603 Pesticides impacts on human health and the environment with their mechanisms of action and 604 possible countermeasures. Heliyon 10:e29128; doi:10.1016/j.heliyon.2024.e29128. 605 Akpolat N, Yahsi S, Godekmerdan A, Yalniz M, Demirbag K. 2005. The value of alpha-SMA in the 606 evaluation of hepatic fibrosis severity in hepatitis B infection and cirrhosis development: a 607 histopathological and immunohistochemical study. Histopathology 47:276–280; doi:10.1111/j.1365-608 2559.2005.02226.x. 609 Arab A, Mostafalou S. 2023. Pesticides and insulin resistance-related metabolic diseases: Evidences 610 and mechanisms. Pesticide Biochemistry and Physiology 195:105521; 611 doi:10.1016/j.pestbp.2023.105521. 612 Arciello M, Gori M, Maggio R, Barbaro B, Tarocchi M, Galli A, et al. 2013. Environmental Pollution: A 613 Tangible Risk for NAFLD Pathogenesis. IJMS 14:22052–22066; doi:10.3390/ijms141122052. 614 Attema B, Janssen AWF, Rijkers D, van Schothorst EM, Hooiveld GJEJ, Kersten S. 2022. Exposure to 615 low-dose perfluorooctanoic acid promotes hepatic steatosis and disrupts the hepatic transcriptome 616 in mice. Molecular Metabolism 66:101602; doi:10.1016/j.molmet.2022.101602. 617 Barrios-Rodríguez R, Pérez-Carrascosa FM, Gómez-Peña C, Mustieles V, Salcedo-Bellido I, Requena P, 618 et al. 2021. Associations of accumulated selected persistent organic pollutants in adipose tissue with 619 insulin sensitivity and risk of incident type-2 diabetes. Environment International 155:106607; 620 doi:10.1016/j.envint.2021.106607. 621 Baudry J, Lelong H, Adriouch S, Julia C, Allès B, Hercberg S, et al. 2018. Association between organic 622 food consumption and metabolic syndrome: cross-sectional results from the NutriNet-Santé study. 623 Eur J Nutr 57:2477–2488; doi:10.1007/s00394-017-1520-1. 624 Baudry J, Rebouillat P, Allès B, Cravedi J-P, Touvier M, Hercberg S, et al. 2021. Estimated dietary 625 exposure to pesticide residues based on organic and conventional data in omnivores, pesco-626 vegetarians, vegetarians and vegans. Food and Chemical Toxicology 153:112179; 627 doi:10.1016/j.fct.2021.112179. 628 Capitão A, Lyssimachou A, Castro LFC, Santos MM. 2017. Obesogens in the aquatic environment: an 629 evolutionary and toxicological perspective. Environment International 106:153–169; 630 doi:10.1016/j.envint.2017.06.003. 631 Castorina R, Bradman A, McKone TE, Barr DB, Harnly ME, Eskenazi B. 2003. Cumulative 632 organophosphate pesticide exposure and risk assessment among pregnant women living in an 633 agricultural community: a case study from the CHAMACOS cohort. Environ Health Perspect 634 111:1640–1648; doi:10.1289/ehp.5887. 635 Cedergreen N. 2014. Quantifying Synergy: A Systematic Review of Mixture Toxicity Studies within 636 Environmental Toxicology. A. Nazir, ed PLoS ONE 9:e96580; doi:10.1371/journal.pone.0096580. 637 Chang J, Li J, Wang H, Wang Y, Guo B, Yin J, et al. 2016. Tissue distribution, metabolism and hepatic 638 tissue injury in Chinese lizards (Eremias argus) after a single oral administration of lambda-639 cyhalothrin. Environmental Pollution 218:965–972; doi:10.1016/j.envpol.2016.08.045. 640 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 21 Chousidis I, Ellero-Simatos S, Payrastre L, Payros G, Lippi Y, Blaquière M, et al. 2025. Bisphenol 641 analogue exposure at low concentrations modifies heart-brain functions and transcriptomics in 642 zebrafish larvae. Journal of Hazardous Materials 498:140004; doi:10.1016/j.jhazmat.2025.140004. 643 Christen V, Crettaz P, Fent K. 2014. Additive and synergistic antiandrogenic activities of mixtures of 644 azol fungicides and vinclozolin. Toxicology and Applied Pharmacology 279:455–466; 645 doi:10.1016/j.taap.2014.06.025. 646 Costes S, Bertrand G, Ravier MA. 2021. Mechanisms of Beta-Cell Apoptosis in Type 2 Diabetes-Prone 647 Situations and Potential Protection by GLP-1-Based Therapies. IJMS 22:5303; 648 doi:10.3390/ijms22105303. 649 de Sousa G, Nawaz A, Cravedi J-P, Rahmani R. 2014. A Concentration Addition Model to Assess 650 Activation of the Pregnane X Receptor (PXR) by Pesticide Mixtures Found in the French Diet. 651 Toxicological Sciences 141:234–243; doi:10.1093/toxsci/kfu120. 652 Di Tommaso P, Chatzou M, Floden EW, Barja PP, Palumbo E, Notredame C. 2017. Nextflow enables 653 reproducible computational workflows. Nat Biotechnol 35:316–319; doi:10.1038/nbt.3820. 654 Dinca V, Docea AO, Drocas AI, Nikolouzakis TK, Stivaktakis PD, Nikitovic D, et al. 2023. A mixture of 13 655 pesticides, contaminants, and food additives below individual NOAELs produces histopathological 656 and organ weight changes in rats. Arch Toxicol 97:1285–1298; doi:10.1007/s00204-023-03455-x. 657 Djekkoun N, Lalau J-D, Bach V, Depeint F, Khorsi-Cauet H. 2021. Chronic oral exposure to pesticides 658 and their consequences on metabolic regulation: role of the microbiota. Eur J Nutr 60:4131–4149; 659 doi:10.1007/s00394-021-02548-6. 660 Docea AO, Gofita E, Goumenou M, Calina D, Rogoveanu O, Varut M, et al. 2018. Six months exposure 661 to a real life mixture of 13 chemicals’ below individual NOAELs induced non monotonic sex-662 dependent biochemical and redox status changes in rats. Food and Chemical Toxicology 115:470–663 481; doi:10.1016/j.fct.2018.03.052. 664 Docea AO, Goumenou M, Calina D, Arsene AL, Dragoi CM, Gofita E, et al. 2019. Adverse and hormetic 665 effects in rats exposed for 12 months to low dose mixture of 13 chemicals: RLRS part III. Toxicology 666 Letters 310:70–91; doi:10.1016/j.toxlet.2019.04.005. 667 European Food Safety Authority (EFSA), Carrasco Cabrera L, Di Piazza G, Dujardin B, Marchese E, 668 Medina Pastor P. 2024. The 2022 European Union report on pesticide residues in food. EFS2 22; 669 doi:10.2903/j.efsa.2024.8753. 670 Expertise Collective Inserm. 2013. Pesticides: Effets sur la santé. Les éditions Inserm. Paris. 671 Expertise Collective Inserm. 2021. Pesticides et effets sur la santé. Nouvelles données. Editions EDP 672 science. 673 Fountoucidou P, Veskoukis AS, Kerasioti E, Docea AO, Taitzoglou IA, Liesivuori J, et al. 2019. A mixture 674 of routinely encountered xenobiotics induces both redox adaptations and perturbations in blood and 675 tissues of rats after a long-term low-dose exposure regimen: The time and dose issue. Toxicology 676 Letters 317:24–44; doi:10.1016/j.toxlet.2019.09.015. 677 Friedman SL, Neuschwander-Tetri BA, Rinella M, Sanyal AJ. 2018. Mechanisms of NAFLD 678 development and therapeutic strategies. Nat Med 24:908–922; doi:10.1038/s41591-018-0104-9. 679 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 22 Fujino C, Watanabe Y, Sanoh S, Nakajima H, Uramaru N, Kojima H, et al. 2019. Activation of PXR, CAR 680 and PPARα by pyrethroid pesticides and the effect of metabolism by rat liver microsomes. Heliyon 681 5:e02466; doi:10.1016/j.heliyon.2019.e02466. 682 Groswald AM, Gripshover TC, Watson WH, Wahlang B, Luo J, Jophlin LL, et al. 2023. Investigating the 683 Acute Metabolic Effects of the N-Methyl Carbamate Insecticide, Methomyl, on Mouse Liver. 684 Metabolites 13:901; doi:10.3390/metabo13080901. 685 He B, Ni Y, Jin Y, Fu Z. 2020. Pesticides-induced energy metabolic disorders. Science of The Total 686 Environment 729:139033; doi:10.1016/j.scitotenv.2020.139033. 687 Hectors TLM, Vanparys C, van der Ven K, Martens GA, Jorens PG, Van Gaal LF, et al. 2011. 688 Environmental pollutants and type 2 diabetes: a review of mechanisms that can disrupt beta cell 689 function. Diabetologia 54:1273–1290; doi:10.1007/s00125-011-2109-5. 690 Hernández AF, Parrón T, Tsatsakis AM, Requena M, Alarcón R, López-Guarnido O. 2013. Toxic effects 691 of pesticide mixtures at a molecular level: Their relevance to human health. Toxicology 307:136–145; 692 doi:10.1016/j.tox.2012.06.009. 693 Hoyeck MP, Matteo G, MacFarlane EM, Perera I, Bruin JE. 2022. Persistent organic pollutants and β-694 cell toxicity: a comprehensive review. American Journal of Physiology-Endocrinology and Metabolism 695 322:E383–E413; doi:10.1152/ajpendo.00358.2021. 696 Jabłońska-Trypuć. 2017. The impact of pesticides on oxidative stress level in human organism and 697 their activity as an endocrine disruptor. Journal of Environmental Science and Health, Part B. 698 Jackson E, Shoemaker R, Larian N, Cassis L. 2017. Adipose Tissue as a Site of Toxin Accumulation. In: 699 Comprehensive Physiology (R. Terjung, ed). Wiley. 1085–1135. 700 Kesse-Guyot E, Baudry J, Assmann KE, Galan P, Hercberg S, Lairon D. 2017. Prospective association 701 between consumption frequency of organic food and body weight change, risk of overweight or 702 obesity: results from the NutriNet-Santé Study. Br J Nutr 117:325–334; 703 doi:10.1017/s0007114517000058. 704 Kesse-Guyot E, Rebouillat P, Payrastre L, Allès B, Fezeu LK, Druesne-Pecollo N, et al. 2020. 705 Prospective association between organic food consumption and the risk of type 2 diabetes: findings 706 from the NutriNet-Santé cohort study. Int J Behav Nutr Phys Act 17; doi:10.1186/s12966-020-01038-707 y. 708 Knebel C, Kebben J, Eberini I, Palazzolo L, Hammer HS, Süssmuth RD, et al. 2018a. Propiconazole is an 709 activator of AHR and causes concentration additive effects with an established AHR ligand. Arch 710 Toxicol 92:3471–3486; doi:10.1007/s00204-018-2321-x. 711 Knebel C, Kebben J, Eberini I, Palazzolo L, Hammer HS, Süssmuth RD, et al. 2018b. Propiconazole is an 712 activator of AHR and causes concentration additive effects with an established AHR ligand. Arch 713 Toxicol 92:3471–3486; doi:10.1007/s00204-018-2321-x. 714 Lee H, Gao Y, Kim JK, Shin S, Choi M, Hwang Y, et al. 2023. Synergetic effects of concurrent chronic 715 exposure to a mixture of OCPs and high-fat diets on type 2 diabetes and beneficial effects of caloric 716 restriction in female zebrafish. Journal of Hazardous Materials 446:130659; 717 doi:10.1016/j.jhazmat.2022.130659. 718 Lee Y-M, Ha C-M, Kim S-A, Thoudam T, Yoon Y-R, Kim D-J, et al. 2017. Low-Dose Persistent Organic 719 Pollutants Impair Insulin Secretory Function of Pancreatic b-Cells: Human and In Vitro Evidence. 66. 720 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 23 Léger T, Balaguer P, Le Hégarat L, Fessard V. 2023. Fate and PPARγ and STATs-driven effects of the 721 mitochondrial complex I inhibitor tebufenpyrad in liver cells revealed with multi-omics. Journal of 722 Hazardous Materials 442:130083; doi:10.1016/j.jhazmat.2022.130083. 723 Leroux P. 2003. Modes d’action des produits phytosanitaires sur les organismes pathogènes des 724 plantes. Comptes Rendus Biologies 326:9–21; doi:10.1016/S1631-0691(03)00005-2. 725 Li J, Li X, Zhang Z, Cheng W, Liu G, Zhao G. 2023. High-Fat Diet Aggravates the Disorder of Glucose 726 Metabolism Caused by Chlorpyrifos Exposure in Experimental Rats. Foods 12:816; 727 doi:10.3390/foods12040816. 728 Li W, Xiao H, Wu H, Pan C, Deng K, Xu X, et al. 2022. Analysis of environmental chemical mixtures and 729 nonalcoholic fatty liver disease: NHANES 1999–2014. Environmental Pollution 311:119915; 730 doi:10.1016/j.envpol.2022.119915. 731 Lichtenstein D, Luckert C, Alarcan J, de Sousa G, Gioutlakis M, Katsanou ES, et al. 2020. An adverse 732 outcome pathway-based approach to assess steatotic mixture effects of hepatotoxic pesticides in 733 vitro. Food and Chemical Toxicology 139:111283; doi:10.1016/j.fct.2020.111283. 734 Lippi Y, Soubès F. 2023. MATRiX: a shiny application for Mining and functional Analysis of 735 TRanscriptomics data.; doi:10.17180/DJ0G-5F33. 736 Lukowicz C, Ellero-Simatos S, Régnier M, Polizzi A, Lasserre F, Montagner A, et al. 2018. Metabolic 737 Effects of a Chronic Dietary Exposure to a Low-Dose Pesticide Cocktail in Mice: Sexual Dimorphism 738 and Role of the Constitutive Androstane Receptor. Environ Health Perspect 126:067007; 739 doi:10.1289/EHP2877. 740 Ma Q, Deng P, Lin M, Yang L, Li L, Guo L, et al. 2021. Long-term bisphenol A exposure exacerbates 741 diet-induced prediabetes via TLR4-dependent hypothalamic inflammation. Journal of Hazardous 742

402:123926; doi:10.1016/j.jhazmat.2020.123926. 743 Mardinoglu A. 2018. Systems biology in hepatology: approaches and applications. 744 Mellor CL, Steinmetz FP, Cronin MTD. 2016. The identification of nuclear receptors associated with 745 hepatic steatosis to develop and extend adverse outcome pathways. Critical Reviews in Toxicology 746 46:138–152; doi:10.3109/10408444.2015.1089471. 747 Mesnage R, Mahmud N, Mein CA, Antoniou MN. 2021. Alterations in small RNA profiles in liver 748 following a subchronic exposure to a low-dose pesticide mixture in Sprague-Dawley rats. Toxicology 749 Letters 353:20–26; doi:10.1016/j.toxlet.2021.10.001. 750 Nichols RG, Rimal B, Hao F, Peters JM, Davenport ER, Patterson AD. 2024. Chlorpyrifos modulates the 751 mouse gut microbiota and metabolic activity. Environment International 192:109022; 752 doi:10.1016/j.envint.2024.109022. 753 Podechard N, Ducheix S, Polizzi A, Lasserre F, Montagner A, Legagneux V, et al. 2018. Dual extraction 754 of mRNA and lipids from a single biological sample. Sci Rep 8:7019; doi:10.1038/s41598-018-25332-755 9. 756 Rajak, S., Raza,z, Tewari, A. 2022. Environmental Toxicants and NAFLD: A Neglected yet Significant 757 Relationship. Dig Dis Sci. 758 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 24 Rebouillat P, Vidal R, Cravedi J-P, Taupier-Letage B, Debrauwer L, Gamet-Payrastre L, et al. 2021. 759 Estimated dietary pesticide exposure from plant-based foods using NMF-derived profiles in a large 760 sample of French adults. Eur J Nutr 60:1475–1488; doi:10.1007/s00394-020-02344-8. 761 Rebouillat P, Vidal R, Cravedi J-P, Taupier-Letage B, Debrauwer L, Gamet-Payrastre L, et al. 2022. 762 Prospective association between dietary pesticide exposure profiles and type 2 diabetes risk in the 763 NutriNet-Santé cohort. Environ Health 21:57; doi:10.1186/s12940-022-00862-y. 764 Rives C, Fougerat A, Ellero-Simatos S, Loiseau N, Guillou H, Gamet-Payrastre L, et al. 2020. Oxidative 765 Stress in NAFLD: Role of Nutrients and Food Contaminants. Biomolecules 10:1702; 766 doi:10.3390/biom10121702. 767 Roustan A, Aye M, De Meo M, Di Giorgio C. 2014. Genotoxicity of mixtures of glyphosate and atrazine 768 and their environmental transformation products before and after photoactivation. Chemosphere 769 108:93–100; doi:10.1016/j.chemosphere.2014.02.079. 770 Schmidt FF, Lichtenstein D, Planatscher H, Mentz A, Kalinowski J, Steinhilber AE, et al. 2021. Pesticide 771 mixture effects on liver protein abundance in HepaRG cells. Toxicology 458:152839; 772 doi:10.1016/j.tox.2021.152839. 773 Sousa S, Rede D, Cruz Fernandes V, Pestana D, Faria G, Delerue-Matos C, et al. 2023. Accumulation of 774 organophosphorus pollutants in adipose tissue of obese women - metabolic alterations. 775 Environmental Research 239:117337; doi:10.1016/j.envres.2023.117337. 776 Tait S, Lori G, Tassinari R, La Rocca C, Maranghi F. 2022. In Vitro Assessment and Toxicological 777 Prioritization of Pesticide Mixtures at Concentrations Derived from Real Exposure in Occupational 778 Scenarios. IJERPH 19:5202; doi:10.3390/ijerph19095202. 779 Velmurugan G, Ramprasath T, Swaminathan K, Mithieux G, Rajendhran J, Dhivakar M, et al. 2017. Gut 780 microbial degradation of organophosphate insecticides-induces glucose intolerance via 781 gluconeogenesis. Genome Biol 18:8; doi:10.1186/s13059-016-1134-6. 782 Vinken M, Knapen D, Vergauwen L, Hengstler JG, Angrish M, Whelan M. 2017. Adverse outcome 783 pathways: a concise introduction for toxicologists. Arch Toxicol 91:3697–3707; doi:10.1007/s00204-784 017-2020-z. 785 Wahlang B, Jin J, Beier JI, Hardesty JE, Daly EF, Schnegelberger RD, et al. 2019. Mechanisms of 786 Environmental Contributions to Fatty Liver Disease. Curr Envir Health Rpt 6:80–94; 787 doi:10.1007/s40572-019-00232-w. 788 Wang B, Tsakiridis EE, Zhang S, Llanos A, Desjardins EM, Yabut JM, et al. 2021. The pesticide 789 chlorpyrifos promotes obesity by inhibiting diet-induced thermogenesis in brown adipose tissue. Nat 790 Commun 12:5163; doi:10.1038/s41467-021-25384-y. 791 Wang R, Yang X, Wang T, Kou R, Liu P, Huang Y, et al. 2023. Synergistic effects on oxidative stress, 792 apoptosis and necrosis resulting from combined toxicity of three commonly used pesticides on 793 HepG2 cells. Ecotoxicology and Environmental Safety 263:115237; 794 doi:10.1016/j.ecoenv.2023.115237. 795 Wang X, Lu Q, Guo J, Ares I, Martínez M, Martínez-Larrañaga M-R, et al. 2022. Oxidative Stress and 796 Metabolism: A Mechanistic Insight for Glyphosate Toxicology. Annu Rev Pharmacol Toxicol 62:617–797 639; doi:10.1146/annurev-pharmtox-020821-111552. 798 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 25 Wei Y, Liu W, Liu J. 2023a. Environmentally relevant exposure to cypermethrin aggravates diet-799 induced diabetic symptoms in mice: The interaction between environmental chemicals and diet. 800 Environment International 178:108090; doi:10.1016/j.envint.2023.108090. 801 Wei Y, Wang L, Liu J. 2023b. The diabetogenic effects of pesticides: Evidence based on 802 epidemiological and toxicological studies. Environmental Pollution 331:121927; 803 doi:10.1016/j.envpol.2023.121927. 804 Xiao X, Sun Q, Kim Y, Yang S-H, Qi W, Kim D, et al. 2018. Exposure to permethrin promotes high fat 805 diet-induced weight gain and insulin resistance in male C57BL/6J mice. Food and Chemical Toxicology 806 111:405–416; doi:10.1016/j.fct.2017.11.047. 807 Yang D, Sun X, Wei X, Zhang B, Fan X, Du H, et al. 2023. Lambda-cyhalothrin induces lipid 808 accumulation in mouse liver is associated with AMPK inactivation. Food and Chemical Toxicology 809 172:113563; doi:10.1016/j.fct.2022.113563. 810 Yang JS, Park Y. 2018. Insecticide Exposure and Development of Nonalcoholic Fatty Liver Disease. J 811 Agric Food Chem 66:10132–10138; doi:10.1021/acs.jafc.8b03177. 812 Zhou Y, Zhou B, Pache L, Chang M, Khodabakhshi AH, Tanaseichuk O, et al. 2019. Metascape provides 813 a biologist-oriented resource for the analysis of systems-level datasets. Nat Commun 10:1523; 814 doi:10.1038/s41467-019-09234-6. 815 816 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 26 Figure legends 817 818 Figure 1: Selection of candidate pesticides. 819 820 Figure 2: Impact of the mixture of 4 pesticides on IHH cell metabolism. 821 (A) Impact of the pesticide mixture on nuclear receptor activation in IHH cells. Relative expression of 822 CYP4A11, CYP2B6, CYP3A4, SREBP1c and CYP1A1 mRNAs in IHH cells exposed to the mixture (ITBL) at 823 30 µM for each pesticide for 24 h (n = 6/condition). (B-E) Impact of the pesticide mixture on gene 824 expression in IHH cells. Data from a microarray experiment performed with IHH cells treated with the 825 pesticide mixture (ITBL) at 30 µM for each pesticide in the mixture for 24 h (n = 6/condition). ( B) 826 Principal component analysis (PCA) score plots of the IHH transcriptomic dataset. ( C) Number of 827 differentially up- and downregulated genes in exposed vs. control untreated cells. (D) Heatmap and 828 hierarchical clustering showing the definition of 2 gene clusters ( p ≤ 0.05 and fold change >1.5). ( E) 829 Pathway and transcription factor enrichment analysis in each cluster. For each sample, the raw data 830 were normalized to the average value of all the samples. (F) Impact of the pesticide mixture on 831 triglyceride content in IHH cells. Triglyceride content in IHH cells treated with the pesticide mixture at 832 30 µM for 10 days (n = 3-4). (G, H) Impact of the pesticide mixture on mitochondrial respiration in IHH 833 cells. (G) Oxygen consumption rate (OCR) profiles of IHH cells treated with the pesticide mixture at 30 834 µM for 24 h (n = 3 -4). (H) Basal respiration, maximal respiration, proton leak and ATP production of 835 IHH cells treated with the pesticide mixture. Data are presented as the mean ± SEM. * Treatment 836 effect, * p < 0.05, ** p < 0.01, *** p < 0.001. (A, F, H) Unpaired parametric T test; (G)Two-way ANOVA 837 multiple comparisons test. 838 839 840 Figure 3: Pesticide mixture exposure does not influence mouse body weight, regardless of diet. 841 (A) Experimental design. Eight-week-old male C57BL6J mice were fed a control diet (CD) or a Western 842 diet (WD) for 5 weeks. Both groups were then divided into 3 subgroups: one fed a diet containing the 843 mixture of 4 pesticides (ITBL) and exposed to the acceptable daily intake (CD - or WD -ADI) of each 844 pesticide; one fed a diet containing the mixture of 4 pesticides (ITBL) and exposed to 10 times the ADI 845 (CD- or WD-10-ADI) of each pesticide; and one not exposed to pesticides (CD or WD) for 20 weeks (n = 846 12 animals per group). (B, C) Body weight in each group from week 0 prior to exposure through 25 847 weeks. The bar graphs show the body weight gain at the end of the experiment. (D-G) Relative 848 subcutaneous (sc) (D, F) and epididymal (pg) (E, G) white adipose tissue weight in each group of mice 849 (n = 12 per group). Data are presented as the mean ± SEM (body weight follow -up (B, C), two-way 850 ANOVA followed by a Tuckey’s post-hoc test; bar graphs (B-G), one-way ANOVA followed by a Tuckey’s 851 post-hoc test). 852 853 Figure 4: Pesticide mixture exposure changes hepatic gene expression in a diet-dependent manner. 854 (A-D) Relative liver weight and plasma alanine aminotransferase (ALT) levels of CD-fed (A, B) and WD-855 fed (C, D) mice in each group (n = 12 per group). (E) Representative histological sections (magnification 856 ×100) of liver stained with hematoxylin and eosin (H&E) and estimated liver steatosis score in each 857 group (n = 12 per group). (F) Hepatic triglyceride content in each group (n = 12 per group). (G-K) Data 858 from an RNA -seq experiment performed with liver samples from each group of mice ( n = 8/group). 859 Principal component analysis (PCA) score plots of liver transcriptomic dataset in CD - (G) and WD-fed 860 mice (H). Number of differentially up- and downregulated genes in CD- and WD-fed mice unexposed 861 vs. exposed to ADI, unexposed vs. exposed to 10ADI or exposed to ADI vs. exposed to 10ADI (n.s., 862 nonsignificant) (I). Hierarchical clustering showing the definition of 2 gene clusters ( p ≤ 0.05 and fold 863 change >1) (J) and pathway and transcription factor enrichment analysis in each cluster (K). Data are 864 presented as the mean ± SEM. * Treatment effect, * p < 0.05, one-way ANOVA followed by Tukey's 865 post-hoc test (A-D, G); Kruskal- Wallis followed by Dunn's post-hoc test (F). 866 867 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 27 Figure 5: Urinary pesticide metabolites and mRNA expression level of genes involved in xenobiotic 868 metabolism. 869 (A-D) Normalized intensities of one metabolite of imazalil ( A), two metabolites of thiabendazole ( B), 870 boscalid and two of its metabolites (C), one metabolite of lambda-cyhalothrin (D) measured by UHPLC-871 HRMS in 24 h urine samples of each group of mice (n = 8 per group). Data are presented as the mean 872 ± SEM. *Treatment effect, * p < 0.05, ** p < 0.01 (Kruskal - Wallis followed by Dunn's post -hoc test); 873 n.d.: non -detected metabolites . (E, F) mRNA expression of hepatic genes involved in xenobiotic 874 metabolism in CD- (E) and WD-fed (F) mice exposed and non-exposed to the pesticide mixture (n = 8 875 per group). Data are presented as the mean ± SEM. *Exposed vs. non-exposed mice, *p < 0.05, **p < 876 0.01 (one-way ANOVA followed by Tuckey’s post-hoc test). 877 878 Figure 6: Pesticide mixture exposure exacerbates WD-induced glucose homeostasis perturbations. 879 (A, E) Oral glucose tolerance test (OGTT) performed after 18 wee ks of pesticide exposure in CD- (A) 880 and WD-fed (E) mice in each group (n = 12 per group) and area under the curve (AUC) representing 881 OGTT results. (B, F) Fasting glycemia. (C, G) Fasting insulinemia. (D, H) HOMAR-IR. (I) Islet number per 882 mm2 of pancreas. (J) Percent of section area occupied by islets. Data are presented as the mean ± SEM. 883 * Exposed vs. non-exposed mice, * p < 0.05, ** p < 0.01; # exposed ADI vs. exposed 10ADI, # p < 0.05, 884 ## p < 0.01 (Two-way ANOVA followed by Tukey's post-hoc test (A, E); One-way Anova followed by a 885 Tuckey’s post-hoc test (bar graphs A-J). 886 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint NutriNet-Santé cohort: Pesticides in cluster 3 with relative mean differences > 1.5 Rebouillat et al. 2021 Pesticides acting through mechanisms in their target organism similar to those implicated in MASLD 3 pesticides banned by the EU Association between pesticide exposure and type 2 diabetes 4 pesticide classes Rebouillat et al. 2022 25 pesticides Acetamiprid Anthraquinone Azadirachtin Azoxystrobin Boscalid Carbendazim Chlorpropham Chlorpyrifos Lambda Cyhalothrin Cypermethrin Cyprodinil Difenoconazole Dimethoate Ometoate Fenhexamid Glyphosate Imazalil Imidacloprid Iprodione Malathion Mathamidophos Profenofos Pyrethins Spinosad Tebuconazole Thiabendazole Azoxystrobin Boscalid Chlorpyrifos Lambda Cyhalothrin Cyprodinil Fenhexamid Imazalil Iprodione Malathion Profenofos Tebuconazole Thiabendazole Azoxystrobin Boscalid Chlorpyrifos Lambda Cyhalothrin Imazalil Malathion Profenofos Tebuconazole Thiabendazole Azoxystrobin Boscalid Lambda Cyhalothrin Imazalil Tebuconazole Thiabendazole Azoxystrobin Boscalid Lambda Cyhalothrin Imazalil Thiabendazole 12 pesticides 9 pesticides 6 pesticides 5 pesticides Figure 1 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint Relative mRNA level CYP4A11 (PPARα) CYP2B6 (CAR) CYP3A4 (PXR) SREBP1c (LXR) CYP1A1 (AhR) A. B. * ***1.5 1.0 0.5 0 5 4 3 2 1 0 2.0 1.5 1.0 0.5 0 1.5 1.0 0.5 0 30 20 10 0 CTL ITBL Dim2 (3.4%) Dim 1 (85.7%) CTL ITBL 0 2000 4000 6000 8000 10000 Differentially expressed genes CTL vs ITBL 42531190 2 1 0 20 40 60 80 100 Cell cycle, mitotic Chromosome organization DNA metabolic process Regulation of cell cycle process 0 5 10 15 20 E2F1 MYC TP53 E2F4 -log10(P) 0 5 10 15 20 Response to nutrient levels Nuclear receptors meta pathway Negative regulation of intracellular signal transduction Response to endoplasmic reticulum stress 0 2 4 6 8 10 SP1 STAT3 TP53 ATF4 Cluster 2 : Genes upregulated by ITBL Cluster 1 : Genes downregulated by ITBL GO enrichment Transcription factor enrichment Color Key and Density Plot Density Row Z-Score 0.8 0.6 0.4 0 -3 1 3 0.2 -1 D. E. F. -log10(P) Count Up Down C. 150 100 50 0 ** G. H. 0.0 0.5 1.0 1.5 2.0 Basal respiration Maximal respiration Proton leak ATP production * ratio 400 300 200 100 0 0 50 100 Oligomycine FCCP Antimycine/ Rotenone 0 30 µM * *** Time (min) * * pmol/min/area µg TG/10⁶ cells CTL ITBL Intra-cellular triglyceride content Oxygen consumption rate * *** Oxygen consumption rate Figure 2 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint 10ADI WATpg A. C57BL/6J 8 week-old n=12/group 5 weeks CD WD 20 weeks CD CD-ADI CD-10ADI WD WD-ADI WD-10ADI Weeks Weeks Body weight (g) 45 40 35 30 25 20 0 Body weight gain (g) 25 20 15 10 5 0 Body weight gain (g) 25 20 15 10 5 WATsc weight WATpg weight WATsc weight WATpg weight WATsc weight/ body weight (%) WATsc weight/ body weight (%) WATpg weight/ body weight (%) WATpg weight/ body weight (%) CD CD-ADI CD-10ADI WD WD-ADI WD-10ADI 8 6 4 2 0 8 6 4 2 0 8 6 4 2 0 8 6 4 2 0 * B. D. n=12/group 1 3 5 7 9 11 13 15 17 21 23 2519 1 3 5 7 9 11 13 15 17 21 23 2519 45 40 35 30 25 20 0 C. E. F. G. Figure 3 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint CD CD-ADI CD- * score CD CD-ADI CD-10ADI 1 2 Fatty acid metabolic process Valine, leucine and isoleucine degradation Energy derivation by oxidation of organic compounds Amino acid catabolic process Sulfur compound metabolic process Fatty acid degradation Protein localization Mitochondrial protein degradation Small molecule biosynthetic process Propanoate metabolism 0 5 10 15 20 25 30 Cluster 1 : Genes upregulated by pesticide mixture 0 1 2 3 Pparα Hdac3 Pparγ -log10(P) Cluster 2 : Genes downregulated by pesticide mixture Ribonucleoprotein complex biogenesis Metabolism of RNA Protein processing in endoplasmic reticulum Parvulin-associated pre-rRNP complex Protein folding Asparagine N-linked glycosylation Protein-RNA complex organization RNA localization tRNA metabolic process Positive regulation of protein localization to chromosome 0 5 10 15 20 25 30 35 40 Snai1 Creb1 0 1 2 -log10(P) Unexposed ADI 10ADI CD WD Liver weight ALT Liver weight ALT Liver weight/body weight (%) U/L 8 6 4 2 0 0 100 200 300 8 6 4 2 0 0 100 200 300 Liver weight/body weight (%) U/L Hepatic triglycerides Steatosis score µg/mg Steatosis score 2 0 1 3 2 1 3 0 60 40 20 0 0 200 400 600 800 µg/mg CD CD-ADI CD-10ADI WD WD-ADI WD-10ADI A. B. E. F. J. K. C. D. Dim2 (7%) Dim1 (19.4%) Dim2 (8.2%) Dim1 (30.4%) 20 0 -20 -50 -25 25 0 50 -25 25 0 50 0 40-40 CD CD-ADI CD-10ADI WD WD-ADI WD-10ADI G. H. Figure 4 I. Steatosis score GO enrichment Transcription factor enrichment Count 500 0 1000 1500 2000 Differentially expressed genes UP DOWN n.s.607 611 904 994 CD vs CD-ADI CD vs CD-10ADI CD-ADI vs CD-10ADI WD vs WD-ADI WD vs WD-10ADI WD-ADI vs WD-10ADI n.s. n.s. n.s. CD-PCA WD-PCA preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint IMZ 1-(2.4-dichlorophenyl)- 2-imidazol-1-ylethanol TBZ 5-hydroxy + sulfate TBZ 5-hydro glucuronide BSC parent BSC 5-hydroxy + sulfate BSC 5-hydroxy glucuronide LCT 3-[(Z)-2-chloro-3,3,3-trifluoroprop-1-enyl] -2,2-dimethylcyclopropane-1 -carboxylic acid glucuronide 6 4 2 0 5 10 15 5 10 15 20 25 5 10 15 20 25 5 10 15 5 15 20 25 10 20 30 40 0 0 000 0 Hepatic detoxification genes Relative mRNA level 0 2 1 3 4 5 6 7 Abcb1aAbcc3Abcc12Cyp1a1Cyp2a4 Cyp2a57b1 Cyp2b10Cyp3a11Cyp4a10Cyp7b1 Fmo1 Gsr Gsta1 Gsta2 Mgst3 Slco1a1Ugt1a6b ***** * * * *** ** * * * *** CD CD-ADI CD-10ADI WD WD-ADI WD-10ADI A. B. C. D. E. 0 2 1 3 4 5 6 7 Abcb1aAbcc3Abcc12Cyp1a1Cyp2a4 Cyp2a57b1 Cyp2b10Cyp3a11Cyp4a10Cyp7b1 Fmo1 Gsr Gsta1 Gsta2 Mgst3 Slco1a1Ugt1a6b x10³ x10³ x10³ x10³ x10³ x10³ x10³ Arbitrary unit Arbitrary unit n.d n.d n.d n.d n.d n.d n.d n.d n.d n.d n.d n.d n.d n.d n.d n.d n.d n.d n.d n.d n.d n.d n.d n.d n.d F. Relative mRNA level Figure 5 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint # # Blood glucose (mg/dL) mg/dL ng/mL HOMAR-IR Islet number /mm² % of islet area OGTT Fasting glycemia Insulinemia HOMAR-IR Islet number Endocrine mass AUC 500 400 300 200 100 0 300 200 100 0 2.0 1.5 1.0 0 0.5 30 20 10 0 6.10⁴ 4.10⁴ 2.10⁴ 0 500 400 300 200 100 0 AUC 6.10⁴ 4.10⁴ 2.10⁴ 0*** Time (min) * **** # # # * * -30 0 15 30 60 90 120 -30 0 15 30 60 90 120 300 200 100 0 2.0 1.5 1.0 0 0.5 *** 30 20 10 0 ** 2.0 1.5 1.0 0 0.5 2.5 2.0 1.5 0 0.5 1.0 * * * CD CD-ADI CD-10ADI WD WD-ADI WD-10ADI A. B. C. D. E. F. ** Time (min) mg/dL ng/mL HOMAR-IR G. H. I. J. Blood glucose (mg/dL) Figure 6 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted February 21, 2026. ; https://doi.org/10.64898/2026.02.18.705565doi: bioRxiv preprint